Experimental and predictive investigation of a vapor compression refrigeration system with thermoelectric subcooling and multi-objective optimization of operating conditions.

Refrigeration systems are essential in daily life. However, increasing cooling demand requires developing more efficient systems with lower environmental footprints. The integration of a thermoelectric cooler with vapor compression refrigeration systems has recently attracted attention as a practical means to improve subcooling and enhance overall performance. The present study experimentally and predictively assesses a VCR system with thermoelectric subcooling (VCR-TECS) for several low-global warming potential (GWP) refrigerants. The VCR system was improved by attaching thermoelectric modules to the condenser outlet, and tests were performed at voltages ranging from 1 to 10 V and water flow rates of 1 to 3 L/min. The introduction of thermoelectric subcooling raised evaporator cooling capacity to 8% and improved the system COP by 5-7% at moderate voltages (4- 6 V). Three statistical models (Random Forest, Gradient Boosting, and Linear Regression) were developed using experimental data and validated. This was the best performance by Gradient Boosting, with an R² of more than 0.996 and an RMSE of 0.02, effectively reproducing the nonlinear trends in performance. The proposed optimization is a score-based multi-objective one that maximizes cooling capacity and COP, and minimizes total power input to determine the optimal operating conditions. Maximum unified scores were obtained by R513A (0.37) and R1234yf (0.36) at 4 V, which is an improvement in the performance of 35-40 % as compared with other refrigerants. In summary, the VCR-TECS structure offers a novel, simple, efficient, and environmentally friendly approach to improve refrigeration performance during low-voltage operation.